Paul Graham Essay Generator

composer.json

{
    "require": {
        "fabpot/goutte": "^3.1",
        "ucsdmath/pdf": "^1.4"
    },
    "autoload": {
        "psr-4": {
            "": ""
        }
    }
}

pgessays.php

<?php
/**
 * Paul Graham Essay Generator
 *
 * Author: Harry G (http://github.com/harrygr)
 * Copyright: Licensed under the GPL-3 (http://www.gnu.org/licenses/gpl-3.0.html)
 *
 * This script scrapes Paul Graham's site and generates a PDF file of all the essays
 *
 * Usage:
 *     - copy the file and composer.json into a directory
 *     - run `composer install` to install the dependencies
 *     - run it from your terminal with `php pgessays.php`
 *
 * A PDF file named "PGs Essays.pdf" will be saved in the current directory
 *
 * To prevent repeated DDoS of Paul's site the script caches the results of the scrape
 * in a sections.json file. Delete this file if you want a fresh copy of the essays (i.e. if some new ones are added).
 *
 * L
 * 
 */

require "vendor/autoload.php";

use Goutte\Client;
use UCSDMath\Pdf\Pdf;

class PgEssayGenerator {

    private $client;
    private $pdf;

    private $main_url = 'http://www.paulgraham.com/articles.html';

    private $sections = [];

    public function __construct(Client $client, Pdf $pdf)
    {
        $this->client = $client;
        $this->pdf = $pdf;
    }

    public function generate()
    {
        $crawler = $this->client->request('GET', $this->main_url);

        if (!file_exists('sections.json')) {
            $crawler->filter('table td[width=435] a')->each(function ($node) {
                $title = $node->text();
                $url = $node->link()->getUri();
                $this->addSection($url, $title);
            });
            file_put_contents('sections.json', json_encode($this->sections));
        } else {
            $this->sections = json_decode(file_get_contents('sections.json'), true);
        }

        $this->buildPdf();

    }

    private function addSection($link, $title)
    {
        echo "Generating section '$title'" . PHP_EOL;
        $crawler = $this->client->request('GET', $link);
        $node = $crawler->filter('table[width=435] td[width=435] font, table[width=374] td[width=374] font');

        if (count($node)) {
            $body = $node->html();
        } else {
            $body = file_get_contents($link);
        }

        $this->sections[] = [
            'title' => $title,
            'body'  => $body,
        ];
    }

    private function buildPdf()
    {
        $this->pdf
            ->initializePageSetup('A4', 'Portrait')
            ->setOutputDestination('F')
            ->setFilename('PGs Essays.pdf');
        $count = count($this->sections);

        foreach ($this->sections as $index => $section) {
            $n = $index + 1;

            echo "Adding section $n of $count: '{$section['title']}' to PDF" . PHP_EOL;
            if (is_string($section['title']) and is_string($section['body'])) {
                $this->pdf->appendPageContent('<h1>' . $section['title'] . '</h1>');
                $this->pdf->appendPageContent($section['body']);
            }
        }

        $this->pdf->render();
    }

}

$generator = new PgEssayGenerator(new Client, new Pdf);

$generator->generate();

Hi,

I had to change line 49 to $crawler->filter('table:nth-child(6) td[width=435] a')->each(function ($node) { to avoid scraping the top table (with 'suggested links').

Also, the PDF library no longer works well with modern php (it loads an old version of mpdf), but I haven't had the chance to look into upgrading that..

Thanks for the script!